I’m the Founder-CEO of a startup I’ve been bootstrapping for 2 years now. I’m proactively looking to avoid a future pivot (when it’ll be too late), by possibly pivoting before our MVP is built.
We prototyped an early stage technology which aims to perfect relevancy between search intents, results shown, and landing page content. This is not an MVP yet, it’s a prototype to validate whether the technology can compute results for millions of intents, while giving users an interface to easily build the intents and results.
Originally the technology was built-for the SEM industry. If you can compute millions of Single Keyword Ad Groups (SKAGs) each containing a perfect relevance between the keyword and ad shown, then you’ve perfected relevance in search advertising. We’ve also tested our technology by integrating it into our Python-based templating system, and we were able to manipulate colours, content styles, words, images etc — all from the intents referred. Since our technology organized groups of intents as groups of SKAGs computed, and since we built a dynamic landing page system, we were able to serve 1 landing page for over 875,000 keywords/intents. With a few rules coded into the landing page, we were able to control how the page should shape-shift itself to best answer the intent of the user.
CTR’s were high, and conversions even higher. The technology works for the SEM industry. And it’s main value is that it significantly speeds up the creation of accounts/campaigns with high-relevant ads and their associated keywords/queries. Teams could then focus on building more business / better optimizing what already works really well.
So I went to SEM communities and started to speak about the technology, and how I’d like to build the MVP as a SaaS for marketers. I began to build a small follower base around the concept, and everyone argues that it’s the next step in SEM. But a few fine gentlemen lifted a few points which got me thinking when it comes to the landing page solution:
1) I’ve built this for our custom Python template system, but each marketer, agency will use their own favourite landing page SaaS tool, programming language, or eCommerce platform (we don’t yet support eCommerce feeds in our intent builder, but can). Some of the landing page solutions (Unbounce, Hubspot, LeadPages) or eCommerce providers (Shopify etc) simply don’t give us enough flexibility to fully integrate our dynamic functions in order to apply rules for the page shape-shifting. There’s a few companies that provide third party solutions for Dynamic Keyword Insertion, but we need more than just inserting keywords, we can shapeshift anything on the page based on the intents referred. And their solutions were just too messy to use, it wouldn’t catch on and IMO would be a hard sell for the market. Perhaps larger clients would see the value, and implement our tech directly into their proprietary templating systems — but it’s once again a hard sell, and there’s a smaller market for those clients.
2) Given the above-mentioned, if we built our own landing page system specifically tailored to the needs of our tech — well we’re now entering the landing page market. So now we’d be competing with 1) AdWords/Bing interface 2) Bid management suppliers (we built proprietary bidding algorithms suited for our tech) 3) Landing page suppliers. It’s gutsy, and yes — we’ve completely re-engineered the idea of building AdWords/Bing accounts/campaigns, their management, and the idea of how website should be built for search-based campaigns. Many in the communities have argued that we’ve innovated the space. But it doesn’t change the fact that if we don’t chose the best product/market fit, the company may not eventually survive on the SEM expert/agency industry alone, because:
a) There’s still a small risk (I doubt, but still exists) that if Google/Bing changes some of their keyword match types, changes limits, it can break down our product’s value/economy. I can only imagine how Google would react if they noticed everyone uploading millions of ads. But realistically, if they were to cut out something significant in their platforms, it would also affect their own revenue stream — so once again, I highly doubt it would happen.
b) Our technology would always be chasing features/changes in the AdWords/Bing landscapes — which isn’t a problem. Just food for thought.
c) Many SEM experts/agencies deal with eCommerce (feeds of products), which we haven’t yet tested our technology-with. It’s very possible, just requires some retrofitting and tooling. We haven’t validated our tech’s ability to work with automated product feeds — and what it would require from eCommerce providers.
d) The acquisition cost of SEM experts/agencies can be expensive. It’ll require for them to embrace our method (which yes is awesome, but also required a radical mindset change as you’re now dealing with controlling for millions in 1 interface). It’ll require for them to embrace our method for building landing pages (integrate our tech server-side into their templating systems, and learn how to use it), or use our landing page builder (web-based if we build one). I’m all for education and sales, I mean if the product is just better, it’s better. But I know that sales isn’t linear, and works in very special ways. Features/benefits only isn’t what drives successful sales. Just food for thought.
e) It’ll require money, which we haven’t even yet considered to raise since we’d like to gain traction with a product which fits the right market first (+ have some traction). So given that we’ve proactively built a poorly-coded prototype to prove the MVP’s full-cycle’s potential, we have an opportunity to build an MVP well tailored to the market it best fits. We have the necessary data, and tools to test and benchmark performance, needs and potentially fitting industries.
As much as I’d love to steer the SEM/text ad PPC market towards a more relevant direction, I’m also trying to be realistic on the realities. I realized that I need to speak to others who have experience in other industries, other than just lead gen, SEM, PPC in order to see where product/market-fit can occur. I’d also like to outline some fresh ideas of the tech’s application:
1) Tailor the tech to be integrated-with/acquired-by the largest landing page / eCommerce SaaS providers (natively). Hubspot, Marketo, Unbounce, LeadPages, Shopify etc would all be viable platforms that can benefit from a deep integration with our technology. For example: Hubspot has a IMO very poor Ad building technology, called Hubspot Ads. Honestly, build 1 ad a time in 2017? C’mon guys. I don’t have the necessary experience and knowledge how such relationships are created, and if the idea of building a product in *hopes* for an acquisition/integration by a large vendor is even possible. I’ll need advice on this matter.
2) Tailor the technology to be integrated-with/acquired-by platforms/brands that want a new tech to either internally use to drive traffic, or offer as a SaaS to their clients. I’m referring to platforms like Yellow Pages, Yelp, Quora, Amazon — any platform/brand which deals with driving search text ad traffic for some kind of need. It doesn’t have to be a network either, it can be an Insurance conglomerate for all I know (just an example). Whoever can value maximizing search + intents.
3) Build a search engine with our technology, and build our own “AdWords/Bing” interface which works specifically for how we build results/ads for intents referred. Marry that with models for landing pages with an easy interface to tailor them for intents, and you’ve now brought a next level of relevance into the search market. I understand that Google is using machine learning to clean and rank their data — but IMO it won’t work for everything, nor every website. If a website isn’t tagged properly, it can’t rank it as well. Another example would be: if Google ranks your car body repair shop, and lists you for a query for “car bumper repair”, it doesn’t mean that the user will convert, or find relevance. Our theory is that users need to see the strings of their query in the ad while having their intent understood/directed. Then once they land on a page, they want to see everything about car bumper repair: images, text, gifs, methods etc. So if the website developer/designer didn’t build all that content (which they usually don’t), Google just won’t perfect that relevancy. Google can only point to a potential match, but it can’t organize content on a landing page for a visitor. Our method marries best of both worlds. We help create relevance between query intents and ads, and then carry that intent to the landing page and make sure that the page shape shifts itself based on the intent referred. Yes it requires some thought effort by building the landing page, but given the right tools — we’ve already prototyped and proven that the method isn’t as difficult as it sounds. It just required a very specific method to build the intent-query related data — which then carries to everything else involved in the landing page creation process.
The application would limitless: it can act as a business directory, it can act as a website directory. And the network would be open for anyone to integrate their own custom website/tech into results fuelled by our engine.
I’d see a potential for integration-with/acquisition-by any search engine, directory-based site (Yellow Pages, Yelp etc) who’d like to bring a new level of relevance / method to listing businesses/websites.
This is just a preliminary idea, and I’m unsure which industries/direction the search engine would take for its first content/clients — but given that we’ve already built our tech for the SEM AdWords/Bing space, we can look into ideas of driving traffic straight from those networks to grow our network/engine given that it provides a heightened-level of relevance between queries and intents.
Our intent/ad building technology can be tailored to be used with generic ad templates, or with a very granular approach. It can also be tailored to be used in a crowdsourced/team form. It just really depends which direction we go for product/market-fit.
Whichever direction we take with the tech, has drastic implications on our stack choice and architecture (which we're currently starting to plan).
The current team in the startup involves:
• A Senior Software Engineer from CGI/Yellow Pages with a Masters in Computer Science. He's the CTO.
• A remote AWS DevOps engineer who worked for eBay/Magento.
• A remote Data Scientist who's worked in the advertising industry.
• Myself, acting as the CEO, lead product architect and designer.
I’m looking to speak with mentors, angels, VC’s, entrepreneurs who understand the current situation and can add valuable insights/advice for my current dilemma.
I think this is quite a complex dilemma, it's not going to be an easy decision! I've been in this position - at the end of the day it's going to come down to what feels like the right fit for your team and your company - you're going to need to decide what purpose you want to fulfill. Although, as someone who's worked with the team at Google, they are very good at what they do and are fully aware of the flaws you've noticed. I'd probably suggest the first option, as it allows you to expand scaleably and have diversification of clients (which has killed many tech startups). It would also allow you to have a closer control over revenue and capitals, and you'd be able to expand your model into a number two-like solution in the end (small business and an enterprise solution).
Feel free to give me a call and I'd be happy to help you work through the problem. Hope this is helpful!